codex-parser

Parse Codex CLI NDJSON logs to extract execution summaries and failure details.

4|1|Updated Jan 19, 2026
One-click install
npx skills add https://github.com/rongarede/skill-snapshots --skill codex-parser
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex-parser
Source: https://github.com/rongarede/skill-snapshots/tree/main/uncategorized/codex-parser
Command: npx skills add https://github.com/rongarede/skill-snapshots --skill codex-parser

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill parses Codex CLI NDJSON logs to quickly reveal the results of executions, reducing the need to manually inspect verbose logs.

Core Features & Use Cases

  • Log overview: generate a concise execution summary including command counts, failures, and token usage from NDJSON logs.
  • Failure tracing: identify failing commands and provide details to accelerate debugging.
  • Change inspection: highlight file changes and notable edits reported in the logs.
  • Mode-based queries: filter logs by batch, search for specific events, or drill into individual entries.

Quick Start

Step 1: Generate the index and manifest for Codex outputs bash <codex-outputs>/scripts/codex-index.sh

Step 2: View an overview cat <codex-outputs>/MANIFEST.md

Step 3: Run targeted inquiries (examples) jq '.logs[] | select(.stats.failed_commands > 0) | {file, failed: .failed_commands_detail}' <codex-outputs>/INDEX.json jq '.logs[] | select(.item.type == "item.completed" and .item.type == "reasoning") | .item.text' <log_file>

Frequently Asked Questions about codex-parser

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I parse Codex CLI NDJSON logs to find failing commands?

You can parse Codex CLI NDJSON logs to find failing commands by generating an index with the codex-index.sh script and querying INDEX.json to extract files with failed_commands greater than zero along with their specific failure details.

What is the best way to get an overview of execution results from NDJSON logs?

Generating a MANIFEST.md file via the codex-index.sh script provides an execution overview from NDJSON logs, summarizing command counts, failures, and token usage without manual inspection of verbose entries.

Can I filter Codex CLI logs by batch or specific events?

Yes, you can filter Codex CLI logs by batch or specific events using mode-based queries with jq to select items by type, such as reasoning or item.completed, and extract their associated text from the log files.

How do I inspect file changes highlighted in Codex CLI outputs?

You inspect file changes highlighted in Codex CLI outputs by parsing the NDJSON logs to reveal notable edits, applying targeted jq queries to extract items reporting file modifications during the execution run.

Do I need jq to debug Codex CLI outputs?

While you can generate an overview using the provided shell script without it, you need jq to run targeted inquiries and extract specific failure details or events when debugging Codex CLI outputs.